Secondary Logo

Journal Logo

A novel application of salivary testosterone in systolic heart failure: relationship with exercise capacity, quality of life and cardiac mechanics

Stout, Martin; Pearce, Keith; Williams, Simon G.

Cardiovascular Endocrinology & Metabolism: March 2015 - Volume 4 - Issue 1 - p 28–38
doi: 10.1097/XCE.0000000000000047
Original articles
Free

Introduction There is an association between heart failure (HF) and testosterone deficiency. Salivary testosterone (ST) has been described as a marker of the free fraction of testosterone in healthy populations and has not commonly been utilized clinically. This study aimed to assess the clinical utility of ST in a HF population with a broad range of testosterone concentrations compared with more traditional serum parameters.

Methods A total of 40 men with HF were recruited. Traditional serum testosterone measures and ST measures were collected on two separate occasions. Patients performed a 6-min walk, underwent echocardiography and completed quality of life questionnaires. Bland–Altman plots were constructed to assess the agreement between ST and free testosterone (FT) levels, and to assess the repeatability between traditional and ST measures. Pearson’s correlation was used to determine the relationship between ST levels, traditional serum testosterone parameters and important health outcomes.

Results There was excellent agreement between ST and FT measures [bias 0.087 nmol/l, SD 0.056 nmol/l, 95% limits of agreement (LoA) of +2 SD 0.104 nmol/l and −2 SD −0.122 nmol/l]. ST and FT measures show excellent reproducibility (bias 0.0137 nmol/l, SD 0.021 nmol/l, 95% LoA +2 SD 0.041 nmol/l and −2 SD −0.041 nmol/l for ST, and bias 0.004 nmol/l, SD ±0.0025 nmol/l, 95% LoA ±2 SD +0.055 nmol/l and −0.046 nmol/l). There are similar levels of moderately strong positive correlation between ST levels, other testosterone fractions and endurance capacity/physical domains of quality of life assessed using the Short Form-36 questionnaire.

Conclusion ST can be used as an alternative marker of traditional testosterone parameters in HF without semi-invasive blood sampling. There is a correlation between exercise capacity and physical quality of life domains with all testosterone fractions.

aManchester Metropolitan University, School of Healthcare Science

bDepartment of Cardiology, North West Heart Centre, University Hospital of South Manchester, Wythenshawe Hospital, Manchester, UK

Correspondence to Martin Stout, PhD, Department of Cardiology, North West Heart Centre, University Hospital South Manchester, Wythenshawe Hospital, Manchester M23 9LT, UK Tel +44 161 2914612; fax: +44 161 2914645; e-mail: martin.stout@uhsm.nhs.uk

Received April 25, 2014

Accepted January 12, 2015

Back to Top | Article Outline

Introduction

Morbidity and mortality from many cardiovascular diseases have declined over recent decades; however, heart failure (HF)-related morbidity and mortality continue to rise 1. Globally, it is estimated that around 15 million individuals suffer from HF, with 400 000 new diagnoses annually in the USA 2. In 2000, HF cost 1.91% of the total UK health-service expenditure 3.

HF is associated with decreased testosterone levels. As such, 25–30% of men with HF aged ∼60 years show biochemical evidence of deficiency 4, and studies report that 79% of men with HF present with testosterone deficiency 5. There is high morbidity and mortality associated with testosterone deficiency in HF, with evidence of further reduction in cardiac output, skeletal muscle mass and exercise capacity 6,7. Studies do report associations between serum total testosterone (TT) levels, cardiovascular haemodynamics and exercise capacity in men with HF 8,9. Reduced TT levels may also be responsible for impaired somatic health in HF 10,11.

Traditionally, diagnosis of testosterone deficiency is based on clinical presentation with laboratory confirmation. TT levels are modulated by illness and perhaps, more importantly, sex hormone-binding globulin (SHBG) levels 12. The majority of circulating testosterone is bound to either SHBG or albumin, with up to ∼3% being unbound, as ‘free testosterone’ (FT). Bioavailable testosterone (BioT) is testosterone bound to albumin, in addition to unbound testosterone in circulation. European guidelines 12 propose values for testosterone deficiency in ageing men but advise caution due to significant variability. This guidance suggests that measuring both FT and BioT levels is essential if TT concentrations are borderline, or when there may be an alteration in SHBG. Older men and those with obesity, diabetes mellitus or chronic disease present with abnormalities in SHBG, rendering TT an imprecise marker of androgen status 13.

Salivary testosterone (ST) is a filtrate of plasma containing only the free fraction of testosterone. Laboratory measurement of serum FT levels is not cost-effective for most clinical laboratories, and estimation from published formulae provides little consensus 14. Therefore, measurement of ST could be a more accurate, convenient and feasible alternative 15,16.

The primary aim of this study was to assess the relationship between ST levels and traditional serum parameters of testosterone in HF patients. In addition, the secondary aims of this study were to assess the relationship between ST levels and measures of exercise capacity, quality of life and cardiac function in the same population.

Back to Top | Article Outline

Methods

Design

Ethical approval was granted by North West 8 Research Ethics Committee and local approval by University Hospital South Manchester Research and Development Directorate. This study conforms to the Declaration of Helsinki, and all participants provided written informed consent.

Forty male HF patients with low (n=20) or normal testosterone levels were recruited between May 2011 and May 2012. According to serum FT analysis, 17 patients were FT deficient (i.e. ≤0.17 nmol/l on the basis of European Endocrine Society Guidelines). No current consensus exists for low testosterone cutoff values for ST nor BioT.

Inclusion criteria were clinically stable HF, evidence of at least moderate impairment of left ventricular (LV) systolic function, reduced exercise tolerance and over 18 years of age. Exclusion criteria were unstable angina, recent acute myocardial infarction, decompensated HF, moderate or severe valvular heart disease, uncontrolled hypertension, orthopaedic or neurological illness limiting the ability to exercise and the presence of a biventricular pacing device/left ventricular assist device in situ.

Back to Top | Article Outline

Six-minute walk test

The 6-min walk is a symptom-limited exercise test designed to allow individuals to achieve maximum effort tolerance and is effective as a one-off measurement of functional capacity in cardiac patients 17. Baseline heart rate and blood pressure, total distance achieved (m), peak heart rate and blood pressure, and peak perceived exertion (Borg Ratings of Perceived Exertion Scale) were recorded. Walking pace was determined by the patient, and rest periods were permitted. No verbal encouragement was given, but a countdown of the remaining walking time was provided.

Back to Top | Article Outline

Echocardiography

Cardiac characteristics were assessed using commercially available ultrasound equipment (GE, Vingmed, Vivid 7, Norway) and ‘offline’ analytical software (GE, Vingmed, EchoPac, Norway) according to current guidelines 18. Structural parameters included the following: biplane left atrial (LA) volume measurement of preatrial contraction volume, minimal LA volume and maximal LA volume, subsequently used to calculate LA active emptying fraction, LA expansion index and LA passive emptying fraction 19. LV end-diastolic cavity dimension and LV end-diastolic length were measured to calculate the sphericity index. The biplane Simpsons method of discs was used to calculate EF. Right atrial (RA) area was measured during systole and right ventricular (RV) area during both systole and diastole to calculate RV fractional area change. Speckle tracking echocardiography (STE) was used to measure the peak strain of the LA contractile, reservoir and conduit periods. STE was also used to measure LV longitudinal, radial and circumferential strain, LV twist and untwist rates and time to peak untwist rates, back rotation rates and time to back rotation rate of the apex and basal LV cavity and LV torsion using LV internal diastolic length. In addition, it was used to measure RV longitudinal strain. Tissue Doppler imaging (TDI) was used to measure LA longitudinal function during ventricular systole, early ventricular diastole and late ventricular diastole/atrial contraction. In addition, TDI parameters were used to assess LV longitudinal systolic velocities (s’) of the medial and lateral walls, together with diastolic properties using the common ratio E/e’. M-mode echocardiography was used to measure tricuspid and mitral annular systolic plane excursion as an adjunct measure of longitudinal systolic function. All echocardiographic measures were averaged over three cardiac cycles when patients were in sinus rhythm and five cardiac cycles with atrial fibrillation.

Back to Top | Article Outline

Blood analysis

All serum blood samples were drawn from the antecubital vein in the supine position between 0800 h and 1200 h midday. Retrospective analysis of data sheets revealed that all blood and salivary samples for testosterone analysis had been taken between the hours of 08:15 and 09:45. Additional samples were taken at the next HF clinic appointment (minimum of 4 weeks between repeated measures, maximum of 4 months). Initial values were used for data analysis. Serum aliquots were prepared for storage at −80°C at the Clinical Chemistry Department of University Hospital South Manchester.

Back to Top | Article Outline

Traditional testosterone measurements

Traditional testosterone (TT) levels were measured from frozen samples using liquid chromatography mass spectroscopy and a previously validated method by accredited clinical scientists 20. The mean recovery was 93% for TT. The standard curve was linear to 50.0 nmol/l; the lower limit of quantification was 0.25 nmol/l and the interassay/intra-assay coefficients were less than 10% for TT (range of 0.3–0.35 nmol/l). FT and BioT levels were calculated using a widely available formula utilizing the SHBG level and albumin concentration 21.

Back to Top | Article Outline

Salivary testosterone

Participants were asked to perform an unstimulated passive drool technique. The participants were asked to rinse out their mouth with plain water, refrain from eating/drinking and not to brush their teeth for 30 min before collection. The patient was instructed to drool through a plastic straw into a collection vial. Samples were collected alongside traditional serum parameters, frozen at −80°C and stored at the Clinical Chemistry Department at University Hospital South Manchester. For analysis, the samples were thawed, mixed and centrifuged. The clear supernatant was then utilized for the analysis. Sample preparation involved a liquid–liquid extraction requiring a 200 μl sample with D5 testosterone as the internal standard and methyl-tert-butyl ether then placed at −80°C. After 1 h at −80°C, the organic layer was transferred and evaporated by heating and gentle nitrogen gas flow, with the resultant residue being reconstituted with a 500 ml/l methanol mobile phase before transferring to a 96-well microtitre plate. Liquid chromatography was performed using a Waters ACQUITY Ultra Performance Liquid Chromatography system (Waters Co., Manchester, UK) and a C18 ACQUITY 1.8 μm HSS T3 column (21×50 mm) maintained constantly at 45°C. The mass spectrometer was a Waters Quattro Premier XE (Waters Co.), set to positive ionization mode. Binary pump mixing of the mobile phases produced a linear gradient that increased from 50 to 90% methanol for 1.5 min. Overall run time was 3.5 min. Testosterone and D5 testosterone coeluted with clean, discrete and identifiable peaks at a retention time of 1.28 min. Intra-assay and interassay coefficients of variation were less than 15%. The mean recoveries from saliva samples at three concentrations were 95.6, 100.3 and 95.8%.

Back to Top | Article Outline

Chronic heart failure classification, depression and quality of life

The New York Heart Association classification was used to grade patients according to functional limitation. Disease/symptom-specific quality of life was assessed using the Minnesota Living with Heart Failure Questionnaire (MLHFQ) 22, and generic health-related quality of life was assessed using the Medical Outcomes Study Short Form (SF-36v2) 23. The Beck Depression Inventory (BDI) was used to measure depression 24.

Back to Top | Article Outline

Power calculation

As this study was not designed to test group differences and was cross-sectional in nature, a power calculation was not performed.

Back to Top | Article Outline

Statistical analysis

Data were analysed using statistical software (SPSS version 19; SPSS Inc., Chicago, Illinois, USA). Data were tested for normal distribution using the Shapiro–Wilk test and were logarithmically transformed, if necessary, before being reassessed for normality before analysis. Bland–Altman plots 25 were used to assess the relationship between ST and serum FT levels and also to assess the reproducibility of ST and serum testosterone repeated measures collected on separate occasions by plotting graphically the bias or mean difference between measures together with the 95% limits of agreement (LoA; equating roughly to ±1.96 SD). Pearson’s product moment correlation was used to assess the relationship between TT measures and ST with important outcome measures in an HF population. The coefficient of determination was calculated for significant results to provide a more complete interpretation of ‘ρ’ and to bring more meaning to the data. A Holm–Bonferroni adjustment 26 was applied to decrease the likelihood of a type II error in highly correlated test statistics. This is considered more powerful than the classical Bonferroni adjustment. Statistical significance was set at a P-value of 0.05 or lower and, where necessary, results were expressed as means±SD.

Back to Top | Article Outline

Results

Patient recruitment summary

A total of 76 patients were identified. Twenty patients (26%) did not reply to their initial invitation letter and eight (11%) replied as ‘not interested’. Outcome measure data were collected at the initial clinic visit, with 100% compliance and no drop-outs. Figure 1 details recruitment and participant flow through the study. Table 1 documents the demographics of the cohort. To summarize, the mean age was 69.4±10.73 years and all participants were male with known severely impaired cardiac function (mean ejection fraction of 28.31±7.07%) and a median New York Heart Association score of II. The main aetiology of HF was ischaemia in 55% of the cohort, with all patients currently receiving optimal medical therapy (100% receiving ACE inhibitors, and >90% receiving diuretics, statins and β-blockers).

Fig. 1

Fig. 1

Table 1

Table 1

The overall mean testosterone levels fall within the normal range (13.08±9.68 nmol/l for TT and 0.202±0.08 nmol/l for FT). When stratified using local laboratory guidelines, 20 patients were found to be testosterone deficient for serum TT and 17 patients for FT.

Back to Top | Article Outline

Agreement between serum free testosterone and salivary testosterone levels

Figure 2 shows the agreement between calculated serum FT measurements and laboratory measurements of ST.

Fig. 2

Fig. 2

The Bland–Altman plot for agreement between ST and FT measures shows a bias of 0.087 nmol/l with an SD of 0.056 nmol/l. This results in a +2 SD of 0.104 nmol/l and a −2 SD of −0.122 nmol/l (95% LoA). The calculated coefficient of variation is 0.12 nmol/l.

Back to Top | Article Outline

Reproducibility of repeated measures of salivary testosterone

Figure 3 details the reproducibility of repeated measurement of ST over the two separate HF clinic visits (minimum of 4 weeks and a maximum of 4 months between repeated measures).

Fig. 3

Fig. 3

The Bland–Altman plot for the agreement between repeated ST measurements demonstrates a bias of 0.0137 nmol/l with an SD of 0.021 nmol/l. This results in a +2 SD of 0.041 nmol/l and −2 SD of −0.041 nmol/l (95% LoA). The calculated coefficient of variation is 0.042 nmol/l.

Back to Top | Article Outline

Reproducibility of repeated measures of serum free testosterone

Figure 4 details the reproducibility of repeated measurements of FT over two separate clinic visits taken in conjunction with ST measurements mentioned previously and with the same timescales (i.e. minimum difference of 4 weeks and maximum of 4 months).

Fig. 4

Fig. 4

The Bland–Altman plot for agreement between repeated measures of FT shows a bias between FT measurements of 0.004 nmol/l (SD±0.0025 nmol/l). Therefore, the mean difference±2  SD or 95% LoA is +0.055 nmol/l and −0.046 nmol/l. The coefficient of repeatability is 0.0050 nmol/l.

Back to Top | Article Outline

Correlation between salivary testosterone and traditional serum testosterone parameters

Table 2 indicates that there is a strong positive correlation between ST and traditionally collected serum measurements of testosterone (TT and BioT, P<0.001). The calculated coefficient of determination between ST and TT is 0.74 and between ST and BioT is 0.63.

Table 2

Table 2

Back to Top | Article Outline

Correlation between salivary testosterone and traditionally collected serum testosterone parameters with exercise capacity

Table 3 shows the Pearson’s correlation between measures of testosterone and 6 min walk exercise capacity.

Table 3

Table 3

Table 3 shows moderately strong correlations between ST, FT and BioT and the overall 6 min walk distance (P<0.001). TT correlates more strongly than the other three parameters (r=0.878, P<0.001). The calculated coefficient of determination for exercise distance and ST is 0.63, FT is 0.56, TT is 0.72 and BioT is 0.36.

Back to Top | Article Outline

Correlation between salivary testosterone and traditionally collected serum testosterone parameters with cardiac functional/structural parameter

Table 4 shows the correlation between collected testosterone parameters and indices of LA structure, function and mechanics.

Table 4

Table 4

There were no significant correlations between ST levels nor any traditional measure of testosterone and detailed indices of LA structure and function in a male HF population.

Table 5 shows the correlation between collected testosterone parameters and indices of LV structure, function and mechanics.

Table 5

Table 5

There were no significant correlations between ST nor any traditional measure of testosterone and detailed indices of LV structure and function.

Table 6 shows the correlation between collected testosterone parameters and indices of right heart structure, function and mechanics.

Table 6

Table 6

There are no significant correlations between ST nor any traditional measure of testosterone and detailed indices of right heart structure and function.

Back to Top | Article Outline

Correlation of salivary testosterone and traditionally collected serum testosterone parameters with quality of life outcomes

Table 7 shows the correlation between collected testosterone parameters and indices of health-specific and disease-specific quality of life.

Table 7

Table 7

ST and FT demonstrate moderate correlations with the general health domain of SF-36 (r=0.402 and 0.466, respectively, P=0.04, following Holm–Bonferroni adjustment). All testosterone fractions apart from BioT correlated strongly and significantly with physical function, BioT demonstrating a more moderate correlation (r=0.597, P<0.001). Moderately strong correlations were observed between all testosterone fractions and the role physical and physical summary domains (all Ps<0.001). The calculated coefficients of determination between ST and FT, and the SF-36 domain of general health were 0.16 and 0.22, respectively. The coefficients of determination between ST, FT, TT and BioT, and SF-36 role physical were 0.49, 0.49, 0.44 and 0.33, SF-36 physical function were 0.65, 0.65, 0.73 and 0.36, and SF-36 physical summary were 0.47, 0.55, 0.55 and 0.50, respectively.

Back to Top | Article Outline

Discussion

This is the first study to assess the relationship between ST levels and traditional serum testosterone levels and the relationship between ST and important health outcomes in men with HF. ST levels correlated strongly and positively (P<0.001) with TT and BioT. Bland–Altman analysis revealed close agreement between ST and FT, reinforcing the concept that ST represents the free fraction of testosterone in HF patients. ST correlated well with many important health outcomes in HF patients, including exercise capacity and important aspects of quality of life.

ST sampling as an alternative to TT sampling is a novel application in HF patients. ST has been established to represent the free fraction of testosterone 14, and exogenous administration of testosterone results in ST increases in parallel with serum TT increases, without significant alteration in SHBG binding capacity 27. Bland–Altman plots between ST and FT confirm that this hypothesis upholds in HF patients. There was a small bias of 0.087 nmol/l between ST and FT values (SD of 0.056 nmol/l), resulting in an excellent calculated coefficient of variation of 0.12 nmol/l. Correlation studies also show that ST levels relate strongly and positively to TT and BioT levels in a HF population, with highly significant ρ-values following adjustment for multiple comparisons (0.861 and 0.792, respectively). In addition to this, basic cost analysis using local healthcare tariffs in the host biochemistry laboratory has shown that the total cost per patient for measurement of ST is around £20 per patient. Serum FT analysis is reported to be three times more costly due to the greater amount of preparation time and staff experience required for its analysis. As such, ST could be considered a more cost-effective measurement, given the close relationship with FT observed in the present study.

TT is reported to be at its lowest before 10:00 a.m. 28. Testosterone levels vary markedly throughout a 24-h period 29,30, with ∼30% of deficient men sampled during the afternoon showing normal testosterone concentrations earlier during the day 28. In this study, testosterone fractions were collected between 08:15 and 09:45 h. For accuracy, participants provided separate samples at their next hospital appointment, which ranged between 4 weeks and 4 months after initial testosterone sampling. Bland–Altman analysis of repeated ST and FT measurements demonstrated excellent reproducibility without deviation from a laboratory low to normal concentration during this time period. There was no appreciable difference in the reproducibility of repeated measurements of FT when compared with ST. This statement is reflected by visualization of graphical data with no formal statistical test performed.

There is a strong positive correlation between testosterone fractions and endurance capacity. TT levels appear to be more correlated with exercise capacity (r=0.848, coefficient of determination 0.72). However, the observed difference between TT and ST and FT is negligible (r=0.792, coefficients of determination 0.63 and 0.751, coefficient of determination 0.56, respectively). BioT shows a weaker relationship with endurance capacity (r=0.598, coefficient of determination 0.36) but remains modest and significant following α-level adjustment. The coefficient of determination is lower (0.36) compared with other testosterone components.

Other research has indicated that TT deficiency detrimentally impacts exercise performance in HF patients, showing significant correlations with peak oxygen uptake 9,31 and 6-min walk distance 8. These findings are consistent with the associations observed in elderly men without clinical evidence of HF 32. Improved exercise capacity in HF patients, associated with higher levels of TT, may be due to altered cardiac function 8,9. This hypothesis has been formulated on the basis that TT is inversely correlated with both LV and RV ejection fraction (EF%) 9, and supplementation has been shown to improve indices of cardiac function in animal and human models 6,32. Our study does not support this hypothesis. It is feasible that our study was underpowered to detect such relationships between testosterone and cardiac function; however, the lack of a trend towards a possible relationship between the data suggests that this is unlikely.

Although studies have clearly reported relationships between testosterone concentration and indices of cardiac function 8,32, there is also ample research evidence to suggest that testosterone does not modulate cardiac mechanics despite improvements in exercise capacity in a similar cohort of patients 33. There are some factors other than sample size that may explain the lack of a relationship observed in this study. All patients in the current study had severely impaired LV systolic function (mean ejection fraction 28.31±7.07%), with the majority (55%) being of ischaemic origin. The pathological nature of ischaemic LV systolic dysfunction (particularly following acute myocardial infarction) may serve to diminish the expected correlation between function and testosterone level. For instance, regional myocardial scarring and fibrosis due to ischaemia rarely show improvement in contractility over time and as such, may negate the expected change in function depending on the testosterone level. In relation to this, many parameters pertaining to cardiac mechanics (e.g. strain imaging, twist, untwist, etc.) would be adversely altered in the presence of regional myocardial scarring.

Another important confounder that may ameliorate possible relationships between cardiac function and testosterone relates to the medical therapy prescribed to the study cohort. Almost all patients were administering angiotensin converting enzyme inhibitor inhibitors, β-blockers, diuretics and statins. There is definitive published evidence showing that statin therapy can adversely alter testosterone concentration, particularly in an aged population 34, and this, together with the combined effects of β-blockade and angiotensin converting enzyme inhibitor inhibition, on cardiac function may diminish the expected relationship between testosterone and cardiac function.

A section of patients were found to be in atrial fibrillation at the time of echocardiography (20%). It is a widely held notion that many diastolic parameters of cardiac function are less accurate with coexistent atrial fibrillation – particularly when associated with dilated left atria and a raised LA pressure. These factors may have also had a bearing on the relationships obtained. To further this concept, the inherent limitations of many of the echocardiographic parameters collected in a fairly obese population, with the possibility of diminished image quality and off-plane measurements (mean BMI 27.02±3.97 kg/m2), may have also adversely affected the testosterone–cardiac function relationship.

In support of the above practical limitations, clinical trial echocardiographic guidance suggests measurement of LV ejection fraction on the basis of LV volumes obtained by the method of discs 35. Limitations arise when the apex is foreshortened, the endocardium is inadequately viewed and there is limitation by reliance on only two LV planes. Longitudinal velocity assessment using PW Doppler can also be limited by a number of factors. PW Doppler assessment of tissue movement is only able to provide information on a specific point of the myocardium determined by sample volume positioning, components perpendicular to the ultrasound beam remain unknown and there is significant angle dependency. TDI velocities may also be influenced by global heart motion, movement of adjacent structures and also blood flow 36. Two-dimensional speckle tracking echocardiography demonstrates technical limitations. In more obese populations with limited images, it is possible that endocardial border tracking may be inaccurate. Two-dimensional speckle tracking echocardiography is also limited in patients with acoustic shadowing or reverberations 36. Tracking software algorithms use a-priori knowledge of ‘normal’ LV function and, as such, there may be errors when assessing regional abnormalities or when assessing neighbouring segments 36.

Testosterone may act at the vascular or skeletal muscle level to promote an increase in walking performance. No vascular nor muscular parameters were measured as part of this study, and future research should aim to address these possibilities. Testosterone levels can impact endothelial function 37, skeletal muscle mass and the local synthesis of growth factors (IGF 1) and of contractile proteins 38. Low testosterone levels can also adversely affect other physiological systems including lung function 39, baroreflex sensitivity and autonomic imbalance 31, levels of circulating inflammatory cytokines found to impede LV contractility and muscular performance 32 and skeletal muscle perfusion 40.

Correlation analysis showed that FT and ST correlate modestly with the general health domain of SF-36 (r=0.402 and.466, respectively). Importantly, the coefficients of determination for this relationship are relatively low at 0.16 and 0.22, respectively, suggesting that this correlation is relatively weak. TT and BioT also demonstrated a modest correlation, which did not reach significance. There were, however, moderately strong, significant correlations between all fractions of testosterone and the physical domains of SF-36, supported by a modest coefficient of determination for the role physical domain (0.49 for ST and FT, 0.44 for TT and 0.33 for BioT). Coefficients of determination for physical function were slightly improved (0.64 for ST, 0.65 for ST, 0.73 for TT), apart from that for BioT, which was modest (0.36). Modest coefficients of determination were observed for the physical summary domain (0.47 ST, 0.55 FT, 0.55 TT and 0.50 BioT). No significant relationships were observed between testosterone and the mental domains of the SF-36 or MLHFQ/BDI scores. Testosterone-deficient elderly men receiving supplementation have been shown to have a significantly improved overall quality of life, as assessed using the SF-12 questionnaire, when compared with those receiving placebo 41. Logically, the significant relationships between exercise capacity and testosterone concentration in this study could translate to improved overall perception of physical and mental quality of life in relation to the overall testosterone level. However, this has not always been the case. Malkin et al. 6 observed no improvement in the general health score following testosterone supplementation in HF patients despite significant improvements in exercise capacity.

During this study, no relationships were seen between BDI or MLHFQ score and testosterone concentration. This is in contradiction to other research, which has shown that low levels of circulating anabolic hormones resulted in increased severity of depressive symptoms in young men with HF 42. The same authors were unable to replicate this improvement in depressive symptoms in an older population, a finding similar to that of this study. Clinically, all participants in our study were stable HF patients. This may, in part, explain the lack of observed correlations with HF-specific quality of life (MLHFQ).

Testosterone deficiency has been associated with the metabolic syndrome and can be predictive of weight gain, central obesity, hypertension, insulin resistance and type II diabetes 43. Research has also showed that in men with HF, testosterone supplementation can improve fasting insulin sensitivity and also reduce body mass 44. It is feasible, that insulin resistance may have promoted decreased exercise capacity and indirectly resulted in alterations to physical related quality of life, due to related deleterious effects on skeletal muscle metabolism 45. Further reports have suggested that insulin resistance may further deteriorate cardiac performance in HF patients, due to cellular disruption of cardiac metabolism 46. Blood samples were not taken for fasting insulin nor glucose assessment. Hence, the impact of insulin resistance on exercise tolerance in our cohort of patients is unknown. This warrants further investigation.

Carrying out multiple correlation statistical analysis during this study, it is plausible that we have elevated the chances of incurring a type I error. To correct for this, we applied simultaneous inference to control the familywise error rate. This technique 26 is more powerful than Bonferroni adjustment and can reduce the type I error rate. Another important facet when analysing correlation studies is to recognize that the ρ-value is a more complete representation of the relationship than the α-value itself. Care must be taken to understand that our statistics represent a relationship between fractions of testosterone and some important health outcomes in a HF model, and that they do not imply a cause and effect. The coefficient of determination was calculated to express data as a percentage. Therefore, the percentage of total variation of variable ‘x’ can be accounted for by variation of variable ‘y’. This technique allows our study to provide a more conservative measure of the relationship between two variables, seldom reported by other researchers in correlation studies.

Back to Top | Article Outline

Conclusion

ST was found to have excellent agreement with FT, together with an excellent level of repeatability, which was consistent with traditional FT serum sampling. We noted a strong positive correlation between ST and serum testosterone levels, which may be modulated by circulating proteins, particularly in an elderly HF population with numerous comorbidities. In addition, basic cost analysis was in favour of ST over FT, given the additional time and experience required for serum FT analysis. There were important relationships between all fractions of testosterone and important health outcomes in a HF population. Future research should aim to build upon this work by attempting to understand the mechanistic physiology behind the observed relationships and to achieve a suitable sample size to ascertain a valid ‘cutoff’ range for ST. One could ascertain that this study shows that a valid cutoff for ST could be equal to FT values given the closeness of their relationship in our HF cohort. ST can be used as a marker of FT in HF patients and provides an opportunity to avoid semi-invasive blood sampling, the inherent inaccuracies of FT calculations from specified formulae and the associated costs of ‘gold-standard’ analysis of FT.

Back to Top | Article Outline

Acknowledgements

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline

References

1. Sytkowski PA, Kannel WB, D’Agostino RB. Changes in risk factors and the decline in mortality from cardiovascular disease. The Framingham Heart Study. N Engl J Med 1990; 322:1635–1641.
2. Eriksson H. Heart failure: a growing public health problem. J Intern Med 1995; 237:135–141.
3. Stewart S, Jenkins A, Buchan S, McGuire A, Capewell S, McMurray JJ. The current cost of heart failure to the National Health Service in the UK. Eur J Heart Fail 2002; 4:361–371.
4. Malkin CJ, Jones TH, Channer KS. Testosterone in chronic heart failure. Front Horm Res 2009; 37:183–196.
5. Güder G, Frantz S, Bauersachs J, Allolio B, Ertl G, Angermann CE, Störk S. Low circulating androgens and mortality risk in heart failure. Heart 2010; 96:504–509.
6. Malkin CJ, Pugh PJ, West JN, van Beek EJ, Jones TH, Channer KS. Testosterone therapy in men with moderate severity heart failure: a double blind randomised placebo controlled trial. Eur Heart J 2008; 27:57–64.
7. Malkin CJ, Pugh PJ, West JN, Van Beek E, Hugh-Jones T, Channer KS. Testosterone therapy in men with moderate severity heart failure: a double blind randomised placebo controlled trial. Eur Heart J 2006; 27:57–64.
8. Bocchi EA, Carvalho V, Guimaraes G. Inverse correlation between testosterone and ventricular ejection fraction, haemodynamics and exercise capacity in heart failure patients with erectile dysfunction. Int Braz J Urol 2008; 34:302–312.
9. Jankowska EA, Filippatos G, Ponikowska B, Borodulin-Nadzieja L, Anker SD, Banasiak W, et al.. Reduction in circulating testosterone relates to exercise capacity in men with chronic heart failure. J Card Fail 2009; 15:442–450.
10. Shores MM, Matsumoto AM, Sloan KL, Kivlahan DR. Low serum testosterone and mortality in male veterans. Arch Intern Med 2006; 166:1660–1665.
11. Jankowska EA, Wegrzynowska K, Superlak M, Nowakowska K, Lazorczyk M, Biel B, et al.. The 12-week progressive quadriceps resistance training improves muscle strength, exercise capacity and quality of life in patients with stable chronic heart failure. Int J Cardiol 2008; 130:36–43.
12. Bhasin S, Jasjua GK, D’Agostino R, Covellio AD, Travison TG. Sex hormone binding globulin, but not testosterone, is associated prospectively and independently with incident metabolic syndrome in men. Diabetes Care 2011; 34:2465–2470.
13. Bhasin S, Zhang A, Coviello A, Jasuja R, Ulloor J, Singh R, et al.. The impact of assay quality and reference ranges on clinical decision making in the diagnosis of androgen disorders. Steroids 2008; 73:1311–1317.
14. Macdonald PR, Owen LJ, Wu FC, Macdowall W, Keevil BG. NATSAL team. A liquid chromatography-tandem mass spectrometry method for salivary testosterone with adult male reference interval determination. Clin Chem 2011; 57:774–775.
15. Bhasin S, Cunningham GR, Hayes FJ, Matsumoto AM, Snyder PJ, Swerdloff RS, Montori VM. Testosterone therapy in men with androgen deficiency syndromes: An Endorine Society Clinical Practise Guideline. J Clin Endocrinol Metab 2010; 95:2536–2559.
16. Rosner W, Auchus RJ, Azziz R, Sluss PM, Raff H. Position statement: utility, limitations, and pitfalls in measuring testosterone: an Endocrine Society position statement. J Clin Endocrinol Metab 2007; 92:405–413.
17. Crapo RO, Casaburi R, Coates AL, Enright PL, Macintyre NR, McKay RT, et al.. Guidelines for methacholine and exercise challenge. Official Statement. Am J Resp Crit Care Med 2002; 166:111–117.
18. Douglas PS, DeCara JM, Devereux RB, Duckworth S, Gardin JM, Jaber WA, et al.. Echocardiographic imaging in clinical trials: American Society of echocardiography standards for echocardiographic core laboratories: endorsed by the American College of Cardiology Foundation. Am Soc Echocardiogr 2009; 22:755–765.
19. Nikitin NP, Witte KK, Thackray SD, Goodge LJ, Clark AL, Cleland JG. Effect of age and sex on left atrial morphology and function. Eur J Echocardiogr 2003; 4:36–42.
20. Thienpont LM, Van Uytfanghe K, Blincko S, Ramsay CS, Xie H, Doss RC, et al.. State-of-the-art of serum testosterone measurement by isotope dilution-liquid chromatography-tandem mass spectrometry. Clin Chem 2008; 54:1290–1297.
21. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84:3666–3672.
22. Rector T, Kubo S, Cohn J. Patient’s self-assessment of their congestive heart failure: content, reliability and validity of a new measure – The Minnesota Living with Heart Failure Questionnaire. Heart Fail 1987; 3:198–208.
23. Ware JE, Kosinski M, Dewey JE. How to score version 2 of the SF-36 Health Survey. Lincoln, RI: Quality Metric Incorporated; 2000.
24. Beck AT, Steer RA, Carbin MG. Psychometric properties of the Beck Depression Inventory. 25 years of evaluation. Clin Psychol Rev 1988; 8:77–100.
25. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307–310.
26. Holm S. A simple sequentially-rejective multiple test procedure. Scand J Stat 1979; 6:67–70.
27. Wang C, Nieschlag E, Swedloff RS, Behre H, Hellstrom WJ, Gooren LJ, et al.. ISA, ISSAM, EAU and ASA recommendations: investigations, treatment and monitoring of late-onset hypogonadism in males. Ageing Male 2009; 12:5–12.
28. Diver MJ, Imtiaz KE, Ahmad AM, Vora JP, Fraser WD. Diurnal rhythms of serum total, free and bioavailable testosterone and of SHBG in middle-aged men compared with those in young men. Clin Endocrinol (Oxf) 2003; 58:710–717.
29. Crawford ED, Barqawi AB, O’Donnell C, Morgentaler A. The association of time of day and serum testosterone concentration in a large screening population. BJU Int 2007; 100:509–513.
30. Spratt DI, O’Dea LS, Schoenfeld D, Butler J, Rao PN, Crowley WF Jr. Neuroendocrine–gonadal axis in men: frequent sampling of LH, FSH, and testosterone. Am J Physiol 1988; 254 (5 Pt 1):E658–E666.
31. Caminiti G, Volterrani M, Iellamo F, Marazzi G, Massaro R, Miceli M, et al.. Effect of long-acting testosterone treatment on functional exercise capacity, skeletal muscle performance, insulin resistance, and baroreflex sensitivity in elderly patients with chronic heart failure a double-blind, placebo-controlled, randomized study. J Am Coll Cardiol 2009; 54:919–927.
32. Zhang YZ, Xing XW, He B, Wang LX. Effects of testosterone on cytokines and left ventricular remodeling following heart failure. Cell Physiol Biochem 2007; 20:847–852.
33. Toma M, McAlister FA, Coglianese EE, Vidi V, Vasaiwala S, Bakal JA, et al.. Testosterone supplementation in heart failure: a meta-analysis. Circ Heart Fail 2012; 5:315–321.
34. Schooling CM, Au Yeung SL, Freeman G, Cowling BJ. The effect of statins on testosterone in men and women, a systematic review and meta-analysis of randomized controlled trials. BMC Med 2013; 11: : 57.
35. Gottdiener JS, Bednarz J, Devereux R, Gardin J, Klein A, Manning WJ, et al.. American Society of Echocardiography recommendations for use of echocardiography in clinical trials. J Am Soc Echocardiogr 2004; 17:1086–1119.
36. Mor-Avi V, Lang RM, Badano LP, Belohlavek M, Cardim NM, Derumeaux G, et al.. Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography. J Am Soc Echocardiogr 2011; 24:277–313.
37. Miller VM, Mulvagh SL. Sax steroids and endothelial function: translating basic science into clinical practise. Trends Pharmacol Sci 2007; 28:263–270.
38. Balagopal P, Rooyackers OE, Adey DB, Ades PA, Nair KS. Effects of aging on in vivo synthesis of skeletal muscle myosin heavy-chain and sarcoplasmic protein in humans. Am J Physiol 1997; 273 (4 Pt 1):E790–E800.
39. Svartberg J, Jorde R, Sundsfjord J, Bonna KH, Barret-Connor E. Seasonal variation of testosterone and wait to hip ratio in men: The TROMSO Study. J Clin Endocrinol Metab 2003; 88:3099–3104.
40. Jones RD, Hugh Jones T, Channer KS. The influence of testosterone upon vascular reactivity. Eur J Endocrinol 2004; 151:29–37.
41. Tong SF, Ng CL, Lee BC, Lee VKM, Khoo EM, Lee EG, Tan HM. Effect of long-acting testosterone treatment on men’s quality of life: a bouble-blind RCT. ISMH World Congress. Abstract:ISMH World Congress, 2010. J Men's Health 2010; 7: 323.
42. Bhattacharya RK, Khera M, Blick G, Kushner H, Nguyen D, Miner MM. Effect of 12 months of testosterone replacement therapy on metabolic syndrome components in hypogonadal men: data from the Testim Registry in the US (TRiUS). BMC Endocr Disord 2011; 11: : 18.
43. Malkin CJ, Jones TH, Channer KS. The effect of testosterone on insulin sensitivity in men with heart failure. Eur J Heart Fail 2007; 9:44–50.
44. Savage DB, Petersen KF, Shulman GI. Mechanisms of insulin resistance in humans and possible links with inflammation. Hypertension 2005; 45:828–833.
45. Ashrafian H, Frenneaux P, Opie LH. Metabolic mechanisms in heart failure. Circulation 2007; 116:434–448.
46. Chang WY, Knochenhauer ES, Bartolucci AA, Azziz R. Phenotypic spectrum of polycystic ovary syndrome: clinical and biochemical characteristics of the three major subgroups. Fertil Steril 2005; 83:1717–1723.
Keywords:

exercise; free testosterone; heart failure; quality of life; salivary testosterone

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.